I’ve just begun working with ESP8266’s and bought a Wemos D1 R2 board and a few barebones ESP-01 modules (along with a USB-to-serial adapter to program them).
Working with the Wemos and the barebones ESP-01 are easy enough - download the chip-specific additional files from github, select the board, and flash away. I’ve got the bog standard LED blinky and toggle-LED-through-WiFi-webserver working fine, so I know that my boards are working fine.
I’ve noticed that the Wemos board displays an access point “ESP_54F108” (snapshot attached) before it even connects to my home WiFi network. Even when the Wemos is connected to my home WiFi nework, it still hosts a parallel AP. As long as I have webserver code on the ESP8266, connecting to the AP hosted by the ESP8266 while the ESP8266 is simultaneously connected to the home WiFi netowrk redirects me to the webserver page. It doesn’t matter how many times I burn new code to the WEMOS board. In fact, it doesn’t even matter what the code is. A blinky with no WiFi header included still creates the AP (just that connecting to the AP and going to 192.168.4.1 results in a "page not found’ error).
Why is the Wemos creating an AP, despite flashing new code onto it?
I checked online, and other users have reported that the Wemos board does infact create an AP using the chip ID in the name, so this doesn’t seem like my problem alone.
I’ve fiddled with the IDE flashing settings (Erase Flash - All Flash), but that does not seem to change anything. This leads me to believe that I am somehow not wiping all the code off the SPI memory chip of the Wemos and that there is some time of residual WiFi manager type code still in the flash. Is there some ‘protected bootloader’ space on the Wemos ESP8266? Is this code built-into the bootloader of the Wemos ESP? Can I burn a new bootloader onto the ESP?